Hi Christian and others, Christian Holm Christensen wrote:
> Kevin, you have any idea how to proceed on this? > > Fons, I send you this report too, so you can say what you think. We > can probably expect more of these kind of errors, since ROOT hasn't been > tested on all `exotic' Linux platforms. I hope Frank answered all that you need regarding hppa? I am no minor arch expert unfortunately. I can provide a list by endianness and bit size -- little-endian, 32-bit: i386, s390, mipsel, arm big-endian, 32-bit: powerpc, armeb [*], m68k, sparc, mips, hppa little-endian, 64-bit: amd64, ia64, alpha big-endian, 64-bit: powerpc64 [*], s390x [*], sparc64 [*] [*] = experimental or not yet fully supported (e.g. there are 64-bit kernels for sparc64 machines but most provided packages have a 32-bit userland) > Oh, and another question, I looked at the amd64 build logs (I know this > is slightly off topic, but anyways), and towards the end it said > > g++ -O2 -pipe -Wall -m64 -fPIC -Iinclude -DR__HAVE_CONFIG -pthread > -I/usr/include/mysql -DUSEPCH -include precompile.h -o > mysql/src/TMySQLServer.o -c mysql/src/TMySQLServer.cxx > In file included from mysql/src/TMySQLServer.cxx:23: > /usr/include/mysql/my_global.h:353:24: error: asm/atomic.h: No such > file or directory > make[1]: *** [mysql/src/TMySQLServer.o] Error 1 > rm utils/src/rootcint_tmp.cxx utils/src/RStl_tmp.cxx > make[1]: Leaving directory `/build/buildd/root-system-5.15.07' > make: *** [build-arch-stamp] Error 2 > > For some reason, the MySQL header includes `asm/atomic.h' which for some > reason isn't available. I realise that this is a problem with MySQL on > that platform, but what to do about it? Looks like this has already been filed as #424276 against libmysqlclient15-dev, so in principle you don't need to do anything but wait until the bug is fixed there. (It is "serious" so should be resolved soon.) > I see ia64 fails because gfortran is not available, and g77 should be > used instead. I guess a Build-Depends like > > gfortran [!ia64] | g77 [ia64] | fortran-compiler > > would do the trick? Not sure what you mean. Checking the ia64 log, the gfortran package (and gfortran-4.1) do get installed, and in the ROOT configure script, it finds gfortran. However, the failing line in the ia64 log is: > g77 -O2 -o hbook/src/hntvar2.o -c hbook/src/hntvar2.f > make[1]: g77: Command not found > make[1]: *** [hbook/src/hntvar2.o] Error 127 and this is preceded, immediately after configure runs, by: > # Add here commands to compile the arch part of the package. > /usr/bin/make ASTEPETAG= XROOTDETAG= UNURANETAG= > make[1]: g77: Command not found > make[1]: Entering directory `/build/buildd/root-system-5.15.07' > make[1]: g77: Command not found > make[1]: g77: Command not found so for some reason it is trying to use g77 *anyway* even though the configure script found gfortran. That is probably what should be fixed (or alternately you could switch it to use g77 on all arches?) > I should also exclude root-plugin-hbook on kfreebsd-amd64. Yes... best regards, -- Kevin B. McCarty <[EMAIL PROTECTED]> Physics Department WWW: http://www.princeton.edu/~kmccarty/ Princeton University GPG: public key ID 4F83C751 Princeton, NJ 08544 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
